@ashxhart I'm in. A practical thought: pre-training from scratch eats compute. A faster 1st win is to take a strong open model, train it on UK data (law, NHS, pub sector) then compress it to run on a laptop, offline. I work with ppl who do model compression & I'm happy to help on that bit.
@pareen The only real Signal of progress in a country is when anyone can safely become a 1st gen profitable entrepreneur without bribing, having to use unethical means, without forcefully working under the soft powers of ruling goons of the society!
#iykyk
@radbackwards You just need to talk and partner with the right people who have their heads down and are building and genuinely adding values to the humanity without the cynosure. If you are ready, let’s talk, we are building a causal world model involving our own humanoids all from India. DM.
There is literally no stability to be found in the AI stack. In the last few months:
1.) I moved all my intense recurring workloads from OpenAI to Anthropic and now back to OpenAI
2.) Used GrokBot as an over lay on top of OpenAI and Anthropic but am now switching this to Muse
3.) liked Instinct for a week until Muse came out
4.) Was using Exa and TinyFish only to now discover this new model called Jev which can save me money on all of the tools listed above
In the matter of days, something that was an “oh shit moment” becomes underwhelming. As a consumer of these tools what a fun time to be alive
Who is the solving the claude, chatgpt multiplayer problem? Comment please…I can’t be sharing summaries to my teammates every single time 😫 and the ‘start from zero context’ problems. And yeah, I am not on slack, nor i want to! So claude tag is a NO!